PII-087 - REDUCTION OF NEXT-GENERATION SEQUENCING FILE SIZE FOR EFFICIENT PHARMACOGENOTYPE EXTRACTION.

H. Kaur¹, R. Ly¹, S. Bray², R. Ratcliff², T. Skaar³, T. Shugg³; ¹Indiana University School of Medicine, ²LifeOmic, ³Indiana University School of Medicine, Division of Clinical Pharmacology.

Background: Comprehensive pharmacogenotype extraction requires the use of Binary Alignment/Map (BAM) files, which require intensive storage resources. The computational pharmacogenoptying tool Aldy, which has been previously validated on clinical whole exome sequencing (WES) and whole genome sequencing (WGS) BAMs, can also extract genotypes from targeted sequencing files. Thus, our objective was to assess the accuracy of Aldy in extracting genotypes for 14 major pharmacogenes from clinical WES and WGS BAMs reduced to only containing pharmacogene regions.

Methods: Clinical germline WES (n=164; sequencing by Ashion Analytics) and WGS (n=100; sequencing by NantOmics) BAMs from solid tumor board patients were reduced to contain only pharmacogene regions using SAMtools. Aldy v4.4 (released 12/14/22) was used to extract genotypes for CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP3A4, CYP3A5, CYP4F2, DPYD, G6PD, NUDT15, SLCO1B1, TPMT, and VKORC1. Aldy results from reduced BAMs were compared to those from corresponding full BAMs and to clinically valid panel-based genotyping results from our institutional pharmacogenetics laboratory.

Results: All Aldy genotype calls from reduced WES and WGS BAMs were concordant with those from full BAMs across the 14 analyzed pharmacogenes. For a WES patient determined to have a CYP3A4*1/*22 genotype by genotyping, Aldy incorrectly called CYP3A4*22/*22 from both reduced and full BAMs. Reducing BAM files to only contain pharmacogene regions decreased mean file size from 9.7 GB to 7.6 MB for WES BAMs and from 107.1 GB to 51.8 MB for WGS BAMs, thus saving 1.6 TB and 10.7 TB worth of storage in our WES and WGS cohorts, respectively.

Conclusion: These findings demonstrate ALDY’s ability to accurately extract pharmacogenotypes from reduced BAM files, decreasing required storage resources by >1,000-fold.